unslothai/unsloth

Evaluation loss becomes constant

Open

#1 067 ouverte le 27 sept. 2024

Voir sur GitHub
 (7 commentaires) (0 réactions) (0 assignés)Python (5 658 forks)batch import
currently fixinggood first issue

Métriques du dépôt

Stars
 (64 271 stars)
Métriques de merge PR
 (Merge moyen 3j 15h) (525 PRs mergées en 30 j)

Description

Hi!

I am testing out unsloth to fine tune llama 3.1 8B instruct and following your notebook here.

One exception is that I have added an eval set. What is really strange is that the eval loss locks up to a specific value after around 300 steps. I mean down to the last decimal, not just flattening out. The training loss looks fine and as expected.

I have changed many parameters and tried different things but it always happens. Any idea on what can cause this?

Guide contributeur